142 Proceedings of Royal Society of Edinburgh. [sess. 
stage it sometimes gets wet in its passage through F. This is 
caused by air rushing in between D and F, and blowing the water 
over it. This wetting can be easily avoided by keeping the stop- 
cock on 0 open, and drawing the stage out slowly ; or, better still, 
by blowing air into the receiver through T before removing the 
stage. The pressure inside drives the water down and keeps the 
stage dry. Perhaps the stage might be more easily taken out dry 
if the tube F were carried up to the height of the stage, so that its 
interior might be always dry. But as this would give rise to dis- 
turbing currents round the stage every time expansion was made, 
on account of the air in the space between the tubes expanding and 
rushing upwards, the advantage therefore seems doubtful. As the 
plan shown works well with care, I have not thought it worth while 
testing the other. 
It will be noticed that the tube K in the centre of the diaphragm 
I, which slides on the centre tube D, has its lower end widened. 
The object of this is, that when D is introduced through the bottom 
of the receiver, its top end shall be guided into its place without 
trouble, and the stage prevented from touching the lower edge 
of K. The diaphragm I forms a floor at the level of the counting 
stage, and is kept up to its correct level by the indiarubber tube M 
pressing it against the stop B. 
Measuring Apparatus. 
In the measuring apparatus described in the previous paper 
there are defects which evidently make it unsuitable for everyday 
work. In devising the new apparatus, besides making it a more 
permanent form, one point specially kept in view was to make the 
arrangements such that the time required for measuring should 
be short, in order that the air might lose as few of its particles as 
possible while it was being measured. Two different plans have 
been tried, and both found to work well. Each of these plans 
possesses certain advantages over the other, and it is for the operator 
to determine which will best suit the conditions under which his 
tests are to be made. The two plans are shown in Plate III. Figs. 
1 to 5 show one arrangement, the other is shown at fig. 6. In the 
first, a cylinder and piston is used as the measure ; in the other, the 
bore in the plug of a stopcock. 
